Application of computer vision in performance test of aircraft engine nozzle

摘要

In order to detect atomization angle and uniformity of aircraft engine nozzle, a new method of computer vision is proposed in this paper. Fuzzy C-means clustering and Hough transform are introduced in spray angle and center offset test. Template self-adaptive image matching algorithms is proposed to detect uniformity. The errors of atomization angle and center offset are less than 0.5 degree. In uniformity test, the error of fuel volume is less than ±2ml. The characteristics of this method are good repeatability and higher precision.